1. Field of the Invention
This invention relates, in general, to detection cells for liquid chromatography detection and more particularly to reference electrode configurations and methods for their use.
2. Description of Related Art
Liquid chromatographic analysis of carbohydrates, amino acids and related compounds has an important place among the tools utilized in biotechnology industry, biochemical research and in clinical laboratories. Use of liquid chromatographic columns in combination with pulsed electrochemical detection in three-electrode detection cells under alkaline conditions makes possible separations of unique selectivity and direct detection without derivatization of separated analytes with unsurpassed sensitivity. Until now, the prevailing approach to amperometric detection in highly alkaline mobile phases uses a gold working electrode, a platinum or a titanium counter electrode, and a reference electrode of the second kind such as a silver-silver chloride electrode, mercury-mercurous chloride electrode, mercury-mercurous sulfate electrode or thallium amalgam-thallous chloride electrode. Electrodes of the second kind are electrodes in which the chemical element (mostly metal) in the solid phase is in equilibrium with its sparingly soluble salt placed in the liquid phase (see, e.g., SAWYER et al., “Experimental electrochemistry for chemists,” 1974, p. 34, Wiley, New York). Over the years, working and counter electrodes have been subject of significant innovations and improvements whereas there were relatively fewer improvements of the reference electrode.
With the increasing importance of capillary chromatography and of hyphenated detection techniques, there is a need to miniaturize the detection cells for carbohydrate, amino acid and related compound analysis. However, miniaturization of cells containing a reference electrode of the second kind is generally very difficult due to the space requirements and surface roughness of liquid junctions and because of general bulkiness of the reference electrode body. There have been attempts to utilize reference electrode of the second kind without a liquid junction. The reference electrode than becomes only a solid phase combination of a metal with a layer of its salt in contact with the test solution (see, e.g., BOHS et al., “The UniJet: A New Electrochemical Detector for Microbore Liquid Chromatography,” Current Separations, 1994, vol. 12, pp. 181-186). One of the serious drawbacks of this approach is a requirement of the presence of a counter anion in the test solution. For example for the Ag/AgCl electrode, a sufficient and constant chloride concentration must be maintained in the test solution. The mandated presence of chloride precludes the use of working electrode which could be contaminated by that anion, for example, gold or platinum. Additionally, the sparingly soluble salt involved in one of the equilibriums of electrode of second kind dissolves much more rapidly in a liquid stream than in the static solution forming the liquid junction. Correspondingly, the electrodes of the second kind are much less durable than the same type of electrodes with a liquid junction.
Silver-silver chloride reference electrodes can undergo a change, usually a positive shift, of reference potential during their exposure to alkaline eluents used in chromatographic carbohydrate and amino acid analysis. This leads to excessive potentials being applied to the working electrodes resulting in a gradually decreasing response and/or in narrowing of range of linearity of calibration plots. In extreme cases, working electrodes can be passivated with a loss of detection sensitivity. Other types of reference electrodes such as, for example, mercury-mercurous-chloride electrodes (calomel electrodes), mercury-mercurous sulfate electrodes, and thallium amalgam-thallous chloride electrodes (Thalamid® electrodes) can be affected by alkaline eluents in a similar fashion and affect the functioning of the working electrodes in the same way. All of the above types of reference electrodes include solutions of conductive salts enclosed in liquid junctions forming the interface between the solid parts of the electrode and the electrolyte on the outside. The liquid junctions are known to exude solutions of conductive salts into the electrode surroundings. If such electrode is positioned in a liquid stream inside a detection cell, the salts from the liquid junction can affect functioning of another detector such as mass spectrometer downstream from the detection cell. If the flow through the detector cell is stopped, such as during interruptions of work during a day or overnight, the ionic content of a liquid junction may even contaminate the surface of the working electrode positioned upstream of the reference electrode affecting its behavior during the subsequent period of use. A gold electrode exposed to a contamination by chloride ions exuded from a liquid junction of a reference electrode can serve as an example. Lastly, the constant diffusion of the ions from the liquid junction predetermines the limited lifetime of a reference electrode of the second kind. Its reference potential can remain in the useful range only as long as the concentration of the conductive salt in the liquid junction remains above a certain minimum value.
Another problem with silver-silver chloride reference electrodes and all other reference electrodes of the second kind is the difficulty to design miniaturized detection cells with minimal dead volume using such electrodes. Minimal dead volume is important for preventing peak broadening and other types of peak shape deformation measured as a loss of chromatographic efficiency. Poor chromatographic efficiency leads among other things to poor separation between peaks and to a corresponding decrease in quantitative precision and accuracy of results based on the measurement of peak areas of separated compounds of interest. Available designs of liquid junctions (roughness, bulkiness) generally make it impossible to minimize dead volume to fulfill the requirements of capillary format of high-performance liquid chromatography.
Yet another disadvantage of conventional reference electrodes of the second kind is that the presence of relatively high concentration of potentially electrode contaminating ions exuding from the reference electrode body is decreasing the number of available options for the relative positioning of electrodes of a three electrode cell. A reference electrode of the second kind always has to be positioned downstream from the working electrode(s). The reference electrodes of the second kind also make serial connections of two or more electrochemical detection cells more difficult for the same reason.
In contrast to reference electrodes of the second kind, reference electrodes of the first kind do not exhibit the same bulkiness of the electrode body and do not require liquid junctions. The reference electrodes of the first kind have a metallic or soluble phase of the electrode in a direct equilibrium with its ion (for example H2 and H+; see, e.g., SAWYER).
So far, only reference electrodes of the first kind that have been realized in liquid chromatographic cells are on the basis of solid electrodes. The reference potential of reported solid electrodes for low dead volume liquid chromatographic cells is dependent on changes of electrolyte composition other than the changes in the concentration of hydronium ions. Such electrodes are called pseudo or quasi reference electrodes, and examples of quasi reference electrodes are disclosed by U.S. Pat. No. 4,404,065 to MATSON or U.S. Pat. No. 5,368,706 to BOWERS et al. Important examples of such solid quasi electrodes are palladium (Pd) or palladium oxide (PdOx) reference electrodes used in electrochemical detection cells for liquid chromatography. In the disclosed example of liquid chromatographic cells the counter electrodes are also made of Pd or PdOx. An important limitation of quasi reference electrodes is their instability of reference potential under the conditions of gradient elution in liquid chromatography. That limitation makes the use of detection cell containing the quasi reference electrodes impossible. The reference potential of PdH electrode, on the other hand, does not change at all or does not change to a degree preventing the use of PdH as reference electrodes under the conditions of liquid chromatographic detection. The only significant change of reference potential of PdH occurs under the conditions of a pH gradient. Such change is predictable and defined by Nernst equation in the same way as for example for glass membrane pH indicator electrodes. For example, pH indicator electrodes can compensate for pH induced changes of response of working electrodes (see, e.g., WELCH et al., “Comparison of Pulsed Coulometric Detection and Potential-Sweep Pulsed Coulometric Detection for Underivatized Amino Acids in Liquid Chromatography,”Anal. Chem., 1989, vol. 61, pp. 555-559). The compensative action of pH sensing reference electrodes such as glass membranes but also PdH electrodes achieves baselines which are unaffected by pH gradients in the form of so called gradient rise or of other similar artifacts. A working electrode with an applied potential referenced to a pH indicator electrode produces a response only in the presence of an electroactive analyte and not as a result of change in pH.
Another example of solid state quasi reference electrode made with Pd or a palladium compound is the Pd/PdI2 electrode disclosed by U.S. Pat. No. 6,572,748 to HERRMANN et al. This reference electrode is formed by depositing a metallic palladium layer by thick film technology. Then, the palladium layer is covered by a suitable palladium compound layer through an electrolytic process or through a precipitation.
Palladium-hydrogen (PdH) electrodes in combination with a suitable reference electrode may be used as pH sensors and/or as proton generation electrodes for aqueous and non-aqueous coulometric titrations. See, e.g., SCHWING et al. “COMPARISON OF DIFFERENT PALLADIUM-HYDROGEN ELECTRODE AS pH INDICATORS,” Analytica Chimica Acta, 1956, vol. 15, pp. 379-388; STOCK et al. “THE PALLADIUM ELECTRODE IN AQUEOUS AND NON-AQUEOUS TITRIMETRY”, Analytica Chimica Acta, 1959, vol. 20, pp. 73-78; and DOBSON, “The PallapHode Electrode System,” Platinum Metals Rev., 1981), vol. 25, no. 2, pp. 72-73.
The palladium-hydrogen electrodes for such purposes are prepared by depositing palladium black upon a platinum, gold, or palladium film. The palladium-hydrogen electrode is then created by adsorption of hydrogen from water electrolysis. During the electrolytic process, the source potential for water electrolysis is not controlled. On the other hand, the electrolytic current and time are monitored and controlled. The electrolytic current is high, as much as 50 mA (see SCHWING). This is at least four orders of magnitude higher than the current values suitable for chromatographic detection cells. The currents of the magnitude common in coulometric titrations would generate excessive noise of the detection signal if applied to a reference electrode of an electrochemical detection cell utilized for chromatographic applications.
Other optimized palladium-hydrogen electrodes for use as pH reference electrode were utilized in potentiometric and voltammetric experiments performed in bulk solutions. See, e.g., GOFFE et al., “Internally charged palladium hydride reference electrode-Part 1: The effect of charging current density on long-term stability,” Medical & Biological Engineering & Computing, 1978, vol. 16, pp. 670-676 (hereinafter “GOFFE”; and KELLY et al., “Internally charged palladium hydride reference electrode: II Automatically controlled palladium hydride electrode,” Medical & Biological Engineering & Computing, 1981, vol. 19, pp. 333-339 (hereinafter “KELLY”). Such palladium-hydride reference electrodes were prepared and maintained by continuous internal charging with a power source. The charging current was monitored and controlled for achieving long-term stability of reference potential. On the other hand, the charging potential was not controlled at all even though it was shown that the applied potential dramatically affected the reference potential of the palladium-hydrogen electrode.
So far, solid state reference electrodes of the palladium-hydrogen type have been described only for use in a bulk solution. See, e.g., FLEISCHMANN et al., “A palladium-hydrogen probe electrode for use as a microreference electrode,” J. Scientific Instruments, 1968, series 2, vol. 1, pp. 667-668 (hereinafter “FLEISCHMANN”); MUNASIRI et al., “Palladium-hydrogen electrodes for coulometric titration analysis of acids and bases,” J. Electroanal. Chem., 1992, vol. 332, pp. 333-337 (hereinafter “MUNASIRI”); “Pd/H2 REFERENCE ELECTRODE”, date unknown, Cormet Testing Systems, Helsinki, Finland (hereinafter “Cormet”). See http://www.cormet.fi/pdf/PDF_Pd—H2.pdf. To date, it does not appear that there has been a single report describing the use of such electrodes in flow through chromatographic detection cells.
In the majority of cases, palladium-hydrogen electrodes are prepared and maintained without using a secondary power source providing a continuous supply of hydrogen to the palladium electrode. For, example, palladium-hydrogen electrodes are prepared by treating palladium with hydrogen generated from a non-continuous electrolytic process (see, e.g., FLEISCHMANN and MUNASIRI) or from a pressurized gas cylinder (see, e.g., Cormet). In one commercial example, Pd/H2 electrode is maintained at a saturated state with hydrogen gas from a cylinder for achieving a stable reference potential (see, e.g., Cormet). Such bubbling of hydrogen gas through the solution not only at the reference electrode but also in the vicinity of working and counter electrodes would cause excessive noise levels in the flow through cells, and is thus completely unacceptable for chromatographic detection cells.
In another example, palladium-hydrogen electrodes are prepared by cycling a Pd wire between negative and positive polarity and maintaining a constant current density (see FLEISCHMANN). During use, the PdH2 electrode was disconnected from the power supply. The palladium-hydrogen electrode thus prepared offered a stable reference potential for up to 24 hours. Then, the palladium-hydrogen electrode needed to be charged with electrolytically generated hydrogen again by cycling it again in acidic solution.
During discontinuous charging, the palladium is converted to various forms of palladium hydride, i.e. α, α+β, β phases. The phases are defined by the ratio of H/Pd (see, e.g., DOBSON et al., “Some Experimental Factors which Govern the Potential of the Palladium Hydride Electrode at 25 to 195° C.,” J. Chem. Soc., Faraday Trans. 1, 1972, vol. 68, pp. 749-763; DOBSON et al., “Plateau Potentials of the α+β Palladium Hydride Electrode at Temperatures between 25 and 195° C.,” J. Chem. Soc., Faraday Trans. 1, 1972, vol. 68, pp. 764-772. Without a continuous supply of hydrogen from the outside, the palladium-hydride phases loose their hydrogen content and convert back to pure palladium. and thus recharging of the reference electrode becomes necessary.
Past attempts to avoid bubble formation at the sensing end of continuously charged PdH electrodes used as pH indicator electrodes led to a creation of the so called “internally charged” electrodes (see, e.g., SCHWING. and GOFFE and Kelly) which are exemplifying in the available literature the format of PdH electrodes with a continuous connection to power and without the need for recharging.
The electrode of this type include a palladium wire continuing from a sealed internal chamber into the bulk of a sample solution. A suitable counter electrode is also contained in the internal chamber. The internal portion of the palladium wire and the counter electrode are immersed in a suitable electrolyte contained in the internal chamber. The continuous supply of hydrogen to the Pd electrode is maintained by a constant current supplied to the pair of electrodes in the internal chamber. However, it must be noted, that this format of continuous recharging of PdH electrodes requires a bulky design of the complete electrode assembly with similar limitations to miniaturization as is the case with a reference electrode of the second kind (e.g., silver/silver chloride, calomel etc). Another drawback of internally charged PdH electrodes is the limitation of their lifetime by the amount of protons obtainable to the Pd electrode in the internal chamber at the range of potentials selected for the internal charging of such electrode. This aspect of internally charged PdH electrodes is similar to the lifetime limitations of electrodes of the second kind based on the limited supply of conductive salt in the liquid junctions of electrodes of the second kind.
To date, internally charged PdH electrodes designed by SCHWING and others have not been adapted for use with low internal volume (1 pL to 1 μL) chromatographic detection cells and neither have been continuously charged palladium-hydrogen electrodes without an internal chamber. The potential usefulness of such electrodes on one hand and the considerable delay of their implementation in chromatographic cells on the other hand indicate that both types of continuously charged PdH electrodes have been deemed unsuitable for use in the low volume flow-through chromatographic detection cell by those skilled in the art until present. It is interesting to observe that similar limitations as with PdH electrodes were encountered with platinum-hydrogen (PtH2) electrodes. See, e.g., GINER, J., “A Practical Reference Electrode,” Electrochem. Soc., 1964, vol. 111, p. 376. The fully functional design of PtH2 with continuous resupply by electrolytic hydrogen was reported already in 1964. However, a format of the same electrode suitable for use in chromatographic flow through detection cells has not yet been reported.
There appear to be formidable obstacles that prevent solid reference electrodes of the first kind for chromatographic three electrode cells of the thin layer type (internal volume 1 pL to 1 μL). First, hydrogen generation at a Pt or Pd substrate requires a presence of a counter electrode, or a second electrode of the reference electrode system, adding to the overall complexity of the detection cell design. Second, bubble formation (e.g., hydrogen gas at cathode, oxygen gas at anode) required for the maintenance of hydrogen electrodes could be considered incompatible with the thin-layer design of known chromatographic detection cells. Third, the power supply circuitry of the reference electrode module has to be integrated into the electronics that is driving the remaining two electrodes and also processing the resulting signal.
Palladium (Pd) has been considered a poor substitute for platinum (Pt) in the metal based hydrogen electrodes because of the lower catalytic activity of Pd for the formation of hydrogen from protons (see, e.g., IVES et al., Reference Electrodes Theory and Practice, 1961, p. 111, Academic Press, New York). Some workers had reported PdH electrodes to be less reliable than Pt-based electrodes (see, e.g., SCHWING).
It would therefore be useful to provide a detection cell with a reference electrode assembly which overcomes the above and other disadvantages of known reference electrodes used in low dead volume (1 pL to 1 μL) electrochemical detection cells.